An adaptive trial design to optimize dose-schedule regimes with delayed outcomes

Area of Science:

• Clinical Trials
• Biostatistics
• Pharmacology

Background:

• Optimizing drug dose-schedule regimens is crucial for balancing toxicity and efficacy.
• Existing clinical trial designs may not adequately address ordered disease subgroups or delayed outcome evaluations.
• Prior biological information can guide the optimization of experimental agents across patient subgroups.

Purpose of the Study:

• To propose a two-stage phase I-II clinical trial design for optimizing dose-schedule regimes.
• To manage the toxicity-efficacy trade-off within ordered disease subgroups.
• To address challenges of delayed outcome evaluation and fast patient accrual in adaptive trial designs.

Main Methods:

• Developed a flexible Bayesian hierarchical model to associate subgroups and regimes.
• Employed a likelihood-based approach treating unobserved outcomes as missing data.
• Utilized elicited utilities and adaptive randomization based on posterior predictive distributions.

Main Results:

• The proposed design effectively optimizes dose-schedule regimes considering toxicity-efficacy trade-offs.
• The Bayesian model accounts for subgroup associations and ordered effects.
• Simulation studies demonstrated the design's performance under various scenarios, including missing data.

Conclusions:

• The novel two-stage adaptive design offers a robust framework for optimizing experimental agent therapy in subgroups.
• This approach provides a practical solution for managing delayed outcomes in clinical trials.
• The design facilitates informed decision-making by integrating toxicity, efficacy, and subgroup information.